Methods and apparatus to identify a mood of media are disclosed. An example method includes accessing a known audio known to evoke a first emotion. A first synthesized sample is synthesized based on the known audio sample. A first value of a first feature of the first synthesized sample is calculated. A mood model is created based on the first feature. The mood model is to establish a relationship between the first feature and the first emotion. A second value of the first feature of first media evoking an unknown emotion is identified. The unknown emotion is identified as the first emotion when the mood model indicates that the second value corresponds to the first value.